The present invention relates generally to moderating an online discussion, and more specifically, to moderating an online discussion based on a graphical text analysis of the expressions made by the discussion participants.
Online discussions take place in online discussion forums, where people hold discussions in the form of posted messages. An online discussion may become unproductive, when “trolls” are about to take charge, when the social dynamics begin to indicate lack of productivity, or when the social dynamics begin to indicate a dysfunctional discussion and a degree of risk, possible retribution, or lawsuit. For some online discussion forums, human monitors disable comments when a discussion becomes unproductive, insulting, tending toward danger, or tending toward bullying or exhibiting threatening dynamics.